Negative doubling: in favor of a big NegP analysis